Heterogeneous delay tomography based on graph fourier transform in mobile networks

Network tomography enables us to monitor communication qualities of network components from end-to-end measurements. We consider a wide area mobile network including heterogeneous network components such as base stations, routers, and servers, and aim to estimate delays at these components. In the heterogeneous network environment, different types of network components have different statistical characteristics of delays. In this paper, we propose heterogeneous delay tomography in mobile networks, which is a new network tomography scheme based on graph Fourier transform (GFT). We assume that average delays of neighboring base stations are comparable and most of the servers have small delays. Under these assumptions, the proposed scheme estimates delays at base stations in the GFT domain and delays at servers with Compressed Sensing. We evaluate the performance of the proposed scheme with simulation experiments.